Presentation | 2018-11-21 Malware Analysis Method Using Conditional Generative Adversarial Network Keisuke Furumoto, Ryoichi Isawa, Takeshi Takahashi, Daisuke Inoue, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Many schemes combining deep learning with methods for imaging malware have been proposed. These methods are considered to be statistical optimization problems unlike conventional manual analysis work. In the data set of malware, the naming rules and criteria of the correct label (malware family) are different for each security vendor, and there is a problem that the identification accuracy is greatly affected by the labeling method. Therefore, we consider approaches to apply Generative Adversarial Networks to malware analysis. In the GAN model, it is possible to learn feature quantities by unsupervised learning. In particular, this paper proposes a method using a model called Conditional GAN. The feature of the proposed method is that it is not necessary to associate correct labels with each specimen in a large-scale dataset, and label information from multiple security vendors can be used in the model. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Malware / Deep Learning / Generative Adversarial Networks |
Paper # | ICSS2018-57 |
Date of Issue | 2018-11-14 (ICSS) |
Conference Information | |
Committee | ICSS |
---|---|
Conference Date | 2018/11/21(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Yoshiaki Shiraishi(Kobe Univ.) |
Vice Chair | Hiroki Takakura(NII) / Katsunari Yoshioka(Yokohama National Univ.) |
Secretary | Hiroki Takakura(NTT) / Katsunari Yoshioka(NICT) |
Assistant | Akira Yamada(KDDI labs.) / Keisuke Kito(Mitsubishi Electric) |
Paper Information | |
Registration To | Technical Committee on Information and Communication System Security |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Malware Analysis Method Using Conditional Generative Adversarial Network |
Sub Title (in English) | |
Keyword(1) | Malware |
Keyword(2) | Deep Learning |
Keyword(3) | Generative Adversarial Networks |
1st Author's Name | Keisuke Furumoto |
1st Author's Affiliation | National Institute of Information and Communications Technology(NICT) |
2nd Author's Name | Ryoichi Isawa |
2nd Author's Affiliation | National Institute of Information and Communications Technology(NICT) |
3rd Author's Name | Takeshi Takahashi |
3rd Author's Affiliation | National Institute of Information and Communications Technology(NICT) |
4th Author's Name | Daisuke Inoue |
4th Author's Affiliation | National Institute of Information and Communications Technology(NICT) |
Date | 2018-11-21 |
Paper # | ICSS2018-57 |
Volume (vol) | vol.118 |
Number (no) | ICSS-315 |
Page | pp.pp.25-30(ICSS), |
#Pages | 6 |
Date of Issue | 2018-11-14 (ICSS) |